基于统计模型的延迟焦化过程关键参数分析与优化

Analysis and Optimization of Key Parameters for Delayed Coking Process Based on Statistical Model

提出了基于统计模型的易操作的延迟焦化关键参数优化方法。首先基于大量的工业装置数据,利用最小二乘法回归出置信度高的关于能耗和产品收率的多项式模型。其次选取操作温度、压力和循环比作为优化变量,以装置能耗和物料效益的权衡为目标函数,考虑建立包含能耗、收率、循环比、加热炉出口温度、操作压力约束在内的多种参数同时优化的非线性模型,并利用建模软件GAMS求解该模型。最后采用边际价值分析,辨识各个参数之间的耦合关系及敏感性,为实际操作调优提供指导。通过该方法计算得到某延迟焦化装置优化的操作温度、压力和循环比分别为495.767℃,0.171MPa和0.498,与现有运行状态相比,模型预测可增加经济效益6.19×106 RMB/a。

A convenient method was established based on statistical models for the key parameters optimization of delayed cokers.The polynomial models were obtained by the least square method about energy consumption and product yields,and the operation temperature,pressure and recycle ratio were selected as the optimization variables.Taking the trade-off between energy consumption and material benefit as the objective function,a nonlinear programming model was built and solved by GAMS.Finally,the marginal value analysis was applied to identify the relation and sensitivity of different parameters,which could provide guidance for practical operating optimization.A case study showed that the optimal operating temperature,pressure and recycle ratio were 495.767℃,0.171 MPa and 0.498,respectively.And compared with the existing state,the optimal model showed that the profit was increased by 6.19×106 RMB/a.